Certain statistical problems especially important in clinical research will be studied. In addition to the development of theory, methods will be implemented and tested in studies in the areas of Clinical Pharmacology, Heart Disease, and Oncology. Estimation from censored observations: In studies of mortality in chronic disease, subjects enrolled late will include some still surviving at the time of analysis. Such "censored" observations require special attention in the analysis. Earlier work of the Principal Investigator and of others will be extended. Study of log-linear models: In many cases, when parameters of interest are related non-linearly to observed quantal variables, as in quantal response assays, the relationship may be expressed in "Log-linear" form. Special cases within the general theory will be studied in order to broaden the range of applications. Models for the conduct of clinical trials: The problem of ethical constraints in the design of clinical trials, and decisions about terminating them, is not well accommodated by standard theory. More appropriate models are proposed and the theory and implications for practice will be studied.